Atrial fibrillation (AF), the leading cause of cardioembolic stroke, accounts for 20% of all acute ischemic stroke (AIS) cases [1]. AF-related AIS is likely to be more debilitating and fatal than other ischemic stroke subtypes [2]. Current guidelines recommend using direct oral anticoagulants (DOACs) for stroke prevention in patients with AF as the first choice over vitamin K antagonists, owing to their comparable efficacies and superior safety profiles [3-5]. However, whether the correlation between anticoagulation activity at onset and large artery occlusions (LAOs) applies to patients taking DOACs remains uncertain. We investigated the impact of the plasma anti-factor Xa concentration on the occurrence of LAOs in patients with AISs receiving factor Xa inhibitors.

We prospectively enrolled consecutive patients with non-valvular AF who developed AISs while taking apixaban, edoxaban, or rivaroxaban and presented within 24 hrs of symptom onset between March 2021 and January 2023. We included Korean patients who had confirmed ingesting DOAC within the last 24 hrs. Patients were excluded if their last DOAC dose was administered over 24 hrs before the index stroke or if the timing of their last DOAC dose was uncertain. AF was diagnosed based on a 12-lead electrocardiogram or 24 hrs of Holter monitoring. Non-valvular AF was defined as AF with no history of rheumatic mitral valve disease, a prosthetic heart valve, or mitral valve reconstruction [4, 5]. LAOs were defined as occlusions at the internal carotid artery, middle cerebral arteries (M1 and M2), basilar artery, anterior cerebral artery (A1), and posterior cerebral artery (P1). The presence of LAOs supplying the vascular territory of the ischemic lesion was determined using magnetic resonance angiography.

Peripheral blood specimens were collected with 3.2% sodium citrate VACUETTE blood collection tubes (Greiner Bio-One GmbH, Frickenhausen, Germany). Blood specimens were processed within 4 hrs of phlebotomy. Upon admission, plasma concentrations of specific DOACs were measured using anti-Xa-based heparin liquid-reagent technology (LRT) chromogenic assays (Hyphen BioMed, Neuville Sur Oise, France) with a Sysmex CS-5100 analyzer (Sysmex Corp., Kobe, Japan). The DOAC activities were categorized as “low” (<100 ng/mL) or “high” (≥100 ng/mL). A plasma DOAC concentration of ≤100 ng/mL was previously recommended as a threshold for using thrombolytics to treat AIS based on theoretical and pharmacological data [6, 7]. This study was approved by the Institutional Review Board of Dong-A University Hospital, Busan, Korea (approval number DAUHIRB-21- 019). The need for informed consent was waived because we used anonymous data and the residual samples after the intended tests were completed. All statistical analyses were performed using IBM SPSS (version 28.0; IBM Corp., Armonk, NY, USA).

The median age of the 66 patients enrolled in this study was 76.0 yrs (interquartile range [IQR], 68.5−82.0), and 41 (62.1%) patients were men. The median National Institute of Health Stroke Scale (NIHSS) score was 3 (IQR, 1.0–8.5). Major artery occlusions were observed in 25 patients (37.9%). DOACs were prescribed for 34 (51.5%) and 32 (48.5%) patients for primary and secondary stroke prevention, respectively. Twenty-eight (42.4%), 28 (42.4%), and 10 patients (15.2%) received apixaban, edoxaban, and rivaroxaban, respectively.

The characteristics of the two groups were compared according to the presence of LAOs (Table 1). The median DOAC plasma concentration on admission was 91.4 ng/mL (IQR, 43.5–178.4). DOAC concentrations were low (<100 ng/mL) in 37 (56.1%) patients and high (≥100 ng/mL) in 29 (43.9%) patients. The DOAC plasma concentrations among patients with LAOs were significantly lower than among patients without LAOs (63.1 ng/mL [IQR, 26.9–105.3] versus 136.6 ng/mL [61.9–200.0]; P=0.023). Multivariable analysis revealed low DOAC plasma concentrations as an independent LAO predictor (odds ratio, 4.86 [95% CI, 1.39–17.0]; P=0.013; Table 2).

Whether DOAC concentrations are stable or variable between and within patients remains poorly understood. Low plasma DOAC concentrations were reportedly strongly associated with stroke severity and the presence of LAOs [6, 8]; however, these reports were obtained from Western countries. The median DOAC concentrations were 20%–25% lower in Asians than in non-Asians, resulting in lower anti-factor Xa activity in Asians [9]. Despite these racial differences, we observed an association between low DOAC blood concentrations (<100 ng/mL) and LAOs in patients receiving DOACs. Thus, monitoring plasma DOAC concentrations may help identify patients prone to severe stroke.

Reduced-dose DOAC refers to the administration of a DOAC at a lower-than-recommended DOAC dose. Theoretically, the use of reduced-dosed DOACs would be associated with lower plasma DOAC concentrations and LAOs in patients with AF. However, our current findings and previous data did not show such results [6], which might reflect the small sample size and low number of plasma DOAC-concentration measurements recorded at different times post-DOAC administration.

Our findings suggest that lower DOAC plasma concentrations are associated with an increased LAO risk in patients with AIS and AF receiving DOACs.